Associations between maternal characteristics, infant sleep, and perceived infant temperament in the first year of life: A longitudinal study

doi:10.21203/rs.3.rs-4277284/v1

Background: Maternal factors, such as her sleep, mood and bonding to her infant, are closely linked to her perceptions of the infant. As mothers tend to be the primary informants on their child’s health and behaviors in research, education and medical settings, a better understanding of the factors that influence maternal perceptions could improve methodologies for assessing infants’wellbeing in the absence of direct measurements. The present study examined the associations between maternal characteristics and perceptions of their infant’s sleep and temperament in the first year of life.

Methods: The participants were 195 mothers of 4- to 8-month-old infants (T0), 76 of whom completed questionnaires six months later (T1). The T0 questionnaires assessed infant sleep, mother’s sleep, postpartum depression, and mother-infant bonding. The T1 questionnaires were the same except for postpartum depression, with the addition of an infant temperament questionnaire.

Results: Despite increased total sleep time and decreased nighttime awakening, mothers’ perceptions of infant sleep problems remained stable across time points. Perceived infant sleep difficulties were primarily associated with concurrent measures of infants’ and mothers’ sleep, whereas perceived difficult temperament was associated with maternal postpartum depression symptoms and bonding difficulty.

Conclusions: Our findings suggest that from mothers’ perspective, sleep problems and difficult temperament are different cognitive constructs related to their child’s behavior and personality. Mothers’ perceptions of their infant sleep behaviors are grounded in temporally proximal infant sleep behaviors, whereas perceptions of the infant’s temperament, particularly negative affect, may reflect more long-term effects of maternal mental health.

It has been proposed that sleep regulation and temperament reflect common underlying central nervous system mechanisms [1], wherein poorer self-regulation is associated with more behavioral problems, including sleep difficulties. A large body of evidence suggests that infants described as having a positive mood, being approachable and sociable are also characterized as sleeping longer at night and/or having fewer sleep problems than infants described as fussy with a negative mood [2–5]. For example, Kelmanson [6] reported that babies who were perceived to have more negative affect were also reported to need more physical intervention (e.g., rocking) to fall asleep, slept too little, were restless during sleep, awoke crying, and had greater difficulty returning to sleep without parental intervention.

Sleep difficulty in infants is more frequently reported by mothers with psychopathology symptoms. Thus, mothers with depression symptoms reported more sleep problems in their infants [7, 8]. Likewise, symptoms of childhood trauma [9] or adverse events [10] have been associated with reports of more difficult infant sleep. These reports may reflect actual infant sleep disturbances. For instance, Bat-Pitault et al. [11] compared one night of polysomnography of infants of mothers with significant depression symptoms to infants of nondepressed mothers a few days after birth and at six months. They found that infants of depressed mothers slept less at night, and these differences increased at six months. Conversely, in a large sample of 1-year-old infants, Halal et al. [12] found that mothers with perinatal depression reported that their infants had more nocturnal awakenings compared with mothers without symptoms, yet no differences in objective sleep measures were observed. It is therefore possible that parents who are depressed or anxious experience more distress from their infants’ sleep patterns, which will be reflected in reports of more problematic sleep, regardless of whether these experiences reflect actual infants’ sleep patterns.

Adding to the problem of separating reported sleep difficulties from the mothers’ emotional state is the relationship between sleep behaviors and temperament. As with sleep, some have reported an association between mothers’ mood and their perceptions of the infant’s disposition [13, 14]. Fernandes et al. [15] reported an association between anxious and depressive symptoms and mothers’ perceptions of infant temperament, which was mediated by parenting stress. However, Whiffen and Gotlib [16] reported that while depressed mothers reported more difficulties with infant care and more feelings of bother than nondepressed mothers did, these negative emotions and cognitions were not reflected in maternal ratings of their infant’s temperament.

However, the relationship between perceived sleep problems and perceived temperament has not been studied extensively. In one study, Halpern et al. [2] objectively assessed both sleep and temperament in 3-month-old infants. They found that objective measures of sleep were more closely associated with behaviorally assessed temperament (by the researchers) than with mothers’ perceptions of the child’s temperament, whereas mothers’ levels of stress were more closely related to their perceptions of temperament. Others have found a weak or no relationship [17] between objective sleep measures and maternal reports of sleep, although the authors noted that sleep-related behaviors, namely, bedtime resistance, total sleep time and nighttime awakenings, may contribute to perceptions of temperament. Thus, a better understanding of the relationship between maternal factors and mothers’ perceptions of infant temperament is warranted.

Another important aspect of maternal emotional state that may have direct bearing on her perceptions of the infant is mother–infant bonding. Parental bonding is defined as the emotions and cognitions experienced by a parent toward the child [18]. A large body of evidence indicates that poor maternal bonding is associated with symptoms of depression, PTSD, anxiety and sleep disturbances [19] and with a perceived difficult temperament [20, 21].

In sum, maternal mood, sleep and bonding may all contribute to mothers’ perceptions of their infant. As parents, and mothers in particular, are the primary informants on their child’s health and behaviors to medical professionals, educators and researchers, a better understanding of the mechanisms that influence maternal perceptions could improve methodologies for assessing infants’ wellbeing in the absence of direct measurement. Moreover, parents’ perceptions of their child influence the child’s perception of themselves [22–24]. Therefore, the aim of this study was to assess the contribution of maternal emotional states to her perceptions of her infant’s sleep problems and temperament while considering the bidirectional influence of perceived sleep patterns and perceived temperament. Thus, in this study, we examined the following questions: [1] How stable are perceptions of infant sleep problems given normative developmental changes in sleep consolidation? [2] What is the relationship between maternal sleep, mood and bonding experience and their perceptions of the infant’s sleep, while controlling for perceived temperament? [3] What is the relationship between maternal sleep, mood, bonding and perceptions of infant sleep and temperament? To address these questions, questionnaires were administered to a community sample of mothers at 6 months and 12 months postpartum. The questionnaires queried mothers regarding their infants’ sleep-related behaviors, temperament at 12 months, and mothers’ emotional state, including mood and sleep.

The aim of the present study was to assess the contribution of maternal emotional states to her perceptions of her infant’s sleep problems and temperament. To achieve this aim, questionnaires were administered to a community sample of mothers at 6 months and 12 months postpartum.

Participants

A community sample of participants was recruited as part of a longitudinal study aimed at investigating factors associated with mother–infant bonding and infant sleep patterns over the first year of life. Here, we report data collected from women who completed two time points. Using convenience sampling, mothers were recruited through social media ads and the snowball method. Two hundred and eight women responded to the advertisements, 195 of whom completed all the questionnaires at T0. Of these, 76 completed both T0 and T1. Eligible participants had full-term healthy infants aged 4-8 months at the time of recruitment (see Table 1 for demographic information). Participants were compensated with vouchers with a total value equivalent to $24.

Demographic questions: Nineteen items regarding sociodemographic information such as maternal and infant age, education, marital status, income, and infant health status.

Infant Sleep Questionnaires: Infant sleep was assessed at both T0 and T1 using two measures: a single item (‘Overall Sleep Problems’), which asked respondents to rate their infants’ sleep on a scale of 1 “not a problem” to 10 “very problematic”; the Infant Sleep Questionnaire [ISQ, 9, 25]. The ISQ includes 30 items reporting on infant sleep quantity and quality. Quantitative items provide continuous measures of bedtime, wake-up time, number, and duration of awakenings per night and amount of sleep during the day in the past week. The qualitative items assess behavioral problems associated with sleep scored on a 3-point scale ranging from “rarely” (zero to one time per week) to “usually” (five to seven times per week), with higher scores representing greater difficulty. These items include bedtime resistance (e.g., “child falls asleep in own bed”, “child needs parent in room to fall asleep”), behavioral aspects of sleep quality (e.g., “child sleeps too little”, “child snores loudly”, “child wakes more than once during the night”), and daytime sleepiness (e.g., “child seems tired”). The Cronbach’s alpha of the full scale, excluding the quantitative items, was acceptable (α=.626) at T0 and satisfactory (α=.727) at T1. The internal consistency of the subscales was acceptable for “sleep problems” (T0 α=.541, T1 α=.606) and “bedtime resistance” (T0 α=.693, T1 α=.662), whereas “daytime sleepiness” was excluded from further analyses due to low internal consistency at both timepoints (T0 α=.145, T1 α=.461).

Infant Temperament: The very short form (VSF) of the revised Infant Behavior Questionnaire [26] was administered at T1. Thirty-seven items of the VSF were selected from the validated translated version of the IBQ-R [27]. The questionnaire yields three global dimensions: (1) Surgency/Extraversion (e.g., “When tossed around playfully how often did the baby laugh?”); (2) Negative Emotionality (e.g., “When placed in an infant seat or car seat, how often did the baby squirm and turn his/her body?”); and (3) Regulatory Capacity⁄Orienting (e.g., “How often during the last week did the baby enjoy being read to?”). In this study, the internal consistencies for the different dimensions were as follows: Surgency/Extraversion, α=0.620; Negative Emotionality, α=0.760; and Regulatory Capacity/Orienting, α=0.768.

As this version of the IBQ-R has not been independently validated in Hebrew and some have questioned the three-factor model structure, [28] a confirmatory factor analysis was conducted and is described in detail in the supplementary materials. A three-factor structure was confirmed and explained 25.4% of the total variance. Factor 1: Displays of negative affect (13 items, 8.0% of total variance, Cronbach’s α=.825), which included items from the Negative Emotionality and Surgency dimensions. Factor 2, Displays of positive affect (19 items, 10.9% of total variance, Cronbach’s α=.729), also included items from the Surgency dimension and the Regulatory Capacity dimension. Factor 3: Displays of fear of strangers (3 items, 6.5% of total variance, Cronbach’s α=.563; see Supplementary Table 1). A similar structure has been reported in non-US cohorts [e.g., 29, 28]

Maternal Sleep: The Pittsburgh Sleep Quality Index [PSQI, 30] was used to assess maternal sleep quality at T0 and T1. The PSQI is a measure of subjective sleep quality comprised of 10 questions about sleep habits over a one-month period. The items are combined into seven components: Sleep Latency, Sleep Duration, Sleep Efficiency, Sleep Disturbances, Subjective Sleep Quality, Medication Use and Daytime Dysfunction. The psychometric properties of the translated version have been reported [31]. A global score above 5 is considered a reliable indicator of clinically significant sleep difficulties. The internal consistency in this study was α = 0.75 at T0 and α = 0.83 at T1.

Mother-Infant Bonding: The Postpartum Bonding Questionnaire [PBQ, 32] was administered at T0 and T1. The original questionnaire contains 25 items that capture the mother’s feelings of closeness and warmth toward the infant (e.g., ״The baby doesn’t seem to be mine”; “I love my baby to bits) and competency as a parent (e.g., “I feel confident when caring for my baby”). The respondents rated their agreement with statements on a 6-point Likert scale ranging from 0 (“always”) to 5 (“never”), with higher scores indicating poorer or more pathological bonding. Thus, the lowest possible score is 0, and the highest possible score for the entire PBQ is 115. For ethical reasons, ‘incipient abuse’ items were not included. The items were translated and back-translated in accordance with Brislin’s [33] guidelines, and the translated version was previously used in several studies with adequate psychometric properties [34, 35]. The internal consistency coefficient at T0 was α=0.898, and that at T1 was α=0.856.

Mothers’ Depression Symptoms: The Edinburgh Postnatal Depression Scale [EPDS, 36, 37] is a 10-item instrument designed to assess depressive symptoms in the postpartum period. In the current study, the Hebrew translation [38] was administered at T0. A cutoff point above 9 has been suggested for screening for clinically significant depressive symptoms, but a cutoff above 12 has greater diagnostic specificity. The measure has been validated in women after childbirth and has demonstrated high internal consistency and validity for detecting major depression in the perinatal period. The internal consistency in this study was 0.87, and nine percent were above the cutoff of 12.

Procedure:

The study was approved by the Institutional Review Board of Tel Aviv Yafo Academic College and was conducted in accordance with The World Medical Association Code of Ethics (Declaration of Helsinki). The questionnaires and data were generated using Qualtrics© 2015 software (Qualtrics, Provo, UT, USA). The data were collected over a two-year period between October 2014 and April 2016. Informed consent was obtained online after a detailed description of the study. The women were then directed to the T0 questionnaires, which included demographic measures, infant and mother’s sleep questionnaires, the EPDS and the PBQ. Those who completed the T0 questionnaires in full were approached again after four months. Women who agreed to respond a second time received a unique link that referred them to T1 questionnaires that included infant and mother’s sleep measures, the PBQ and the IBQ. The mean interval between T0 and T1 was 5.7 months (±1.1).

Analytical methods

Statistical analyses were conducted using SPSS® V28.0 (IBM) for initial data cleaning, reliability testing and creation of summary measures. Stata V18 (https://www.stata.com/) was used for regression analyses. Differences between T0 and T1 assessments were calculated using paired t-tests or Wilcoxon signed-rank tests when distributions were not normal. Zero-order Pearson correlations were used to determine associations between the different variables. Hierarchical regressions were used to determine the relative contribution of maternal and infant sleep factors to perceived sleep problems and perceived temperament at T1, followed by Seemingly Unrelated Estimation Test (SUEST) which allows testing the relative contribution of the independent variable to different outcome measures in a single model.

We first examined whether mothers who dropped out differed from the remaining group on demographic measures and mothers’ sleep (PSQI), bonding (PBQ) and postpartum depression symptoms (EPDS) collected at T0. T-tests were used to compare continuous variables between groups, and chi-square tests were used for categorical variables. The cohorts differed with respect to infant age (t₍₁₉₃₎=5.01, p<.001), as infants in the cohort who dropped out were approximately one month younger (M=5.5, SD=1.2) than infants in the cohort who completed both timepoints (M=6.4, SD=0.9). Additionally, a significantly greater percentage of women who dropped out had additional children (51% vs 35%, 𝜒²=7.86, p=.020). As neither of these parameters correlated with measures of interest in the study, we concluded that dropout was likely driven by women’s availability to complete the survey rather than confounding characteristics. The remaining analyses were conducted with the sample that participated at both timepoints.

To answer the first question regarding the stability of mothers’ perceptions of their infant sleep patterns and problems, paired t-tests or Wilcoxon signed-rank tests were used to assess the difference between timepoints, and zero-order Pearson correlations were used to assess agreement of measurements across timepoints of infant sleep variables (Table 2). Mothers’ sleep, measured with the PSQI, was highly correlated across timepoints, with 61% of mothers reporting significant sleep disturbances at T0 compared with 48.9% at T1, with a decrease in sleep disturbance nearing significance (T₍₇₅₎=1.91, p=.060). Bonding, measured with the PBQ, was also highly correlated across timepoints, with no change across the time points (T₍₇₅₎=0.743, p=.459). There were significant changes in the quantitative measures of infant sleep, indicating greater consolidation of sleep during the night. Mothers reported more total sleep time (TST) at night (Wilcoxon=451.0, p<.001), less wake-time after sleep onset (WASO, Wilcoxon=1285.5, p=.002), fewer awakenings (Wilcoxon=1409.0, p=.049), and less sleep during the day (Wilcoxon =1240.0, p<.001). The correlation coefficients were medium to high, with no correlation between T0 and T1 in WASO. At T1, mothers reported less bedtime resistance (T₍₇₅₎=2.59, p=.011) but more nighttime problems (Wilcoxon=680.0, p=.016), and there was no change in the single item for overall sleep problems (Wilcoxon=898.0, p=.102). The correlation coefficients were also intermediate to high.

Zero-order Pearson’s correlation analyses were performed across and within timepoints (Supplementary Table 2). The number of other children and income correlated with each other and with several of the other scales in the study. Therefore, the number of other children was entered as a covariate in subsequent analyses. Correlation coefficients ranged from ‘weak’ to ‘moderate’ strength, with stronger correlations for concurrent assessments and weaker correlations across timepoints. Several observations emerged from the correlation analyses: mothers’ depression symptoms (EPDS) mothers’ sleep (PSQI) and various qualitative and quantitative items from the ISQ (infant sleep) measures correlated with each other both within and across timepoints and were therefore included in hierarchical regression models. Bonding (PBQ) correlated primarily with other concurrent measures, as well with the EPDS and PSQI at T0. To limit the number of dependent variables in hierarchical regression analyses and to maintain a ratio of 10 respondents per variable, PBQ at T0 was not included in further analyses.

To answer the third question regarding the long-term association between maternal sleep, mood and bonding experience with her perceptions of the infant’s sleep and temperament, we calculated the associations of maternal factors at T0 and T1 with the Overall sleep problems item and negative affect factor (Factor 1) of the IBQ using two hierarchical regression analyses, followed by the SUEST to compare the relative contributions of the different maternal and infant factors to perceived sleep and affect difficulties in a single model. In the first hierarchical model, the Overall Sleep Problems item at T1 was the dependent variable, Step 1 included the number of other kids, infant total sleep time at T1, and number of nighttime arousals at T1 as control factors, Step 2 included the PSQI and PBQ at T1, and Step 3 included the PSQI and the EPDS at T0 (Table 3). All three models were significant (step 1: F_(3,68)=9.96, p<.001; step 2: F_(5,66)=9.40, p<.001; step 3: F_(7,64)=7.31, p<.001), with step 2 significantly increasing the explained variance from 30.5% to 42.6% (p=.003). Step 3 further increased the explained variance to 44.4%, but the change was not significant (p=0.204). The factors that accounted for most of the variance in the Perceived Sleep Problems item were the factors associated with infant sleep patterns (amount and number of nighttime arousals), with some contribution from maternal sleep problems (e.g., PSQI).

A similar analysis was run with IBQ Factor 1 (Table 4). As can be seen, steps 2 (F_(5,66)=2.53, p=.037) and 3 (F_(7,64)=2.61, p=.020) were significant, while step 1 was not (F_(3,68)=2.01, p=.121). Step 2 significantly increased the explained variance from 8.3% to 16.3% (p=.050), and step 3 increased the explained variance to 22.5%, but the change was not significant (p=0.090). The factors that accounted for most of the variance in Perceived Sleep Problems item were the number of other kids, bonding difficulty at T1 and depression symptoms at T0.

Finally, to compare the relative contributions of maternal factors and infant sleep to perceptions of both sleep problems and difficult temperaments, the SUEST was run on the third step of both models. The SUEST procedure combines information from models and tests the null hypothesis that the contributions of the different predictors are equivalent across the two models. In both models, having other children significantly contributed to the variance in the dependent factor, with opposite effects (𝜒²=14.84, p<.001). There were no other overlapping significant independent variables, indicating that the variance in perceived sleep problems and the variance in perceived negative affect are explained by different factors (see Figure 1).

Parents are the primary source of information regarding their child’s wellbeing and personality, informing professional caregivers, researchers, and the child herself. The association between maternal emotional states and her perceptions of her infant’s temperament and sleep has been a lingering question [39]. In this study, we examined the stability of maternal perceptions of infant sleep problems given normative developmental changes in sleep patterns; the relationship between maternal sleep, mood and bonding experience; and mothers’ perceptions of the infant’s sleep and temperament. The main findings were that perceptions of infant sleep problems at 12 months were best explained by concurrent infants’ sleep behaviors and mothers’ sleep quality. In contrast, the infant’s temperament, specifically negative affect, was best predicted by maternal depression symptoms six months earlier and more difficulty with bonding to the child. Thus, our findings suggest that from the mother’s perspective, sleep problems and difficult temperament are different cognitive constructs related to a child’s behavior and personality.

In this study, we used the ISQ, one of several parent-reported instruments considered reliable and valid measures of infant sleep patterns [reviewed in 40]. The advantage of the ISQ is that it combines quantitative information regarding infants’ sleep behaviors with subjective maternal experiences. Across time points, we found that perceptions of infant sleep problems were stable even as mothers reported an increase in total sleep time and fewer nighttime awakenings. In contrast to several studies in high-risk populations [14, 41, 42] in which maternal symptoms of depression predicted perceived infant sleep problems, the strongest predictors in the present study were concurrent quantitative aspects of the infant’s sleep and mothers’ sleep quality. Moreover, the strongest correlates of the Overall Sleep Problems item at both six and twelve months were the concurrent measure of nighttime awakenings. These correlations potentially indicate that current difficulty managing the child’s sleep contributes more to perceptions of sleep problems than maternal factors.

Regarding infant temperament, of the three factors that emerged from the IBQ factor analysis, only Factor 1 - Negative Affect - correlated consistently with other measures in the study. The strongest predictors were the number of other kids and maternal depressive symptoms assessed six months earlier. Thus, perceived temperament was better explained by maternal factors than by infant sleep behaviors. Our findings are consistent with Halpern et al, [2] who found the mothers’ perceptions of their child’s temperament were related to levels of stress rather than to objective measures of infant sleep. Nevertheless, others have reported associations between infant sleep and perceived temperament [e.g., 4], although these reports did not control for quantitative measures of infant sleep.

Our findings regarding perceived infant temperament are consistent with observations that maternal depressive symptoms are related to infant irritability and distress, both in studies in which temperament predicted maternal mood symptoms [e.g., 43] and vice versa, when maternal symptoms predicted infant negative, but not positive, aspects of infant temperament [44]. Given that in the present study maternal mood was assessed six months prior to temperament assessment, maternal mood may have long-term effects on perceptions of the infant’s temperament. The caveat here is that maternal mood was not assessed concurrently at twelve months, and it is likely that women with more depression symptoms at six months had residual depression symptoms six months later. It is worth noting that the bonding measure (PBQ), which is reliably linked to postpartum depression [32], was also a significant predictor of IBQ Factor 1 when depression symptoms were not included in the model, suggesting that the shared variance between bonding difficulty and perceived infant temperament at T1 may be explained by depression symptoms.

Regarding bonding, previous studies have found similar associations between reports of difficult temperament and problems with bonding and have interpreted these associations to demonstrate the effects of the infant’s temperament on maternal bonding [43]. Here, bonding, while stable across timepoints, only correlated with perceived temperament concurrently and not predictively. Moreover, as noted above, in the regression model, this association disappeared when the EPDS was entered. These findings add to new evidence suggesting that bonding may be shaped by infant temperament [20, 21]. Future studies are necessary to identify the association between bonding and temperament using time-lagged models.

This study has several important limitations. First, despite the good initial sample size, there was a significant, albeit unbiased, dropout that hindered our ability to apply more complex predictive modeling techniques and compromised the statistical power of the study. Second, only maternal report instruments were used, limiting external validity and the scope of the interpretation vis-à-vis objective infant behaviors. Finally, the sample was a low-risk, middle class, college educated and ethnically homogeneous community sample, limiting the generalizability of the results. This merits replication of the study in more diverse samples with younger infants.

The abovementioned limitations notwithstanding, this study provides several important insights. First, the longitudinal design provided the opportunity to demonstrate stability and change in relationships between maternal characteristics and perceptions of infant sleep and temperament. Specifically, we suggest that mothers’ perceptions of their infant’s sleep behaviors are more strongly linked to their nighttime interactions (assessed by the number of awakenings) with the infant than to their own mood, quality of sleep and bonding experience. In contrast, perceptions of the infant’s temperament, especially expressions of negative affect and frustration, are more strongly linked to maternal psychological wellbeing. Second, this is one of only a few studies to report associations between maternal characteristics, including their sleep and bonding, with maternal perceptions of their infant’s sleep and temperament. More longitudinal studies with larger and more diverse samples are needed to further elucidate the associations between maternal well-being, their perceptions of the infant, and infant outcomes.

Ethics approval and consent to participate: Before completing the questionnaires, informed consent was obtained from all respondents. Failure to endorse the consent section terminated the questionnaire. The study was approved by the Institutional Review Board of the authors’ institutions and was conducted in accordance with The World Medical Association Code of Ethics (Declaration of Helsinki).

Availability of data and materials: Data will be provided upon request.

Competing interests: The authors declare that they have no competing interests.

Funding: This work was funded by Tel Aviv Yafo Academic College Internal Grants.

Author contributions: I. S. H. designed the study and wrote the funding application and manuscript; E. M. was involved in some of the data analyses; and J. U. H. cowrote the funding application.

Acknowledgments: The authors also thank Tal Solnik-Menilo, Dana Deviri and Tamar Lehman-Inbar for their hard work on the project. Additional thanks to Orr-Stav Communications for assistance with the translation of the instruments.

Dahl RE. Sleep, learning, and the developing brain: early-to-bed as a healthy and wise choice for school aged children. Sleep. 2005;28:1498–9.
Halpern LF, Anders TF, Coll CG, Hua J. Infant temperament: Is there a relation to sleep-wake states and maternal nighttime behavior? Infant Behav Dev. 1994;17:255–63.
Scher A, Epstein R, Sadeh A, Tirosh E, Lavie P. Toddlers’ sleep and temperament: reporting bias or a valid link? A research note. J Child Psychol Psychiatry. 1992;33:1249–54.
Sorondo BM, Reeb-Sutherland BC. Associations between infant temperament, maternal stress, and infants’ sleep across the first year of life. Infant Behav Dev. 2015;39:131–5.
Spruyt K, Aitken RJ, So K, Charlton M, Adamson TM, Horne RS. Relationship between sleep/wake patterns, temperament and overall development in term infants over the first year of life. Early Hum Dev. 2008;84:289–96.
Kelmanson IA. Temperament and sleep characteristics in two-month-old infants. Sleep Hypn. 2004;6:78–84.
Hoffman C, Dunn DM, Njoroge WF. Impact of postpartum mental illness upon infant development. Curr Psychiatry Rep. 2017;19:100.
Slomian J, Honvo G, Emonts P, Reginster J-Y, Bruyère O. Consequences of maternal postpartum depression: A systematic review of maternal and infant outcomes. Womens Health. 2019;15:1745506519844044.
Hairston IS, Waxler E, Seng JS, Fezzey AG, Rosenblum KL, Muzik M. The Role of Infant Sleep in Intergenerational Transmission of Trauma. SLEEP. 2011. https://doi.org/10.5665/sleep.1282.
Ciciolla L, Addante S, Quigley A, Erato G, Fields K. Infant sleep and negative reactivity: The role of maternal adversity and perinatal sleep. Infant Behav Dev. 2022;66:101664.
Bat-Pitault F, Sesso G, Deruelle C, Flori S, Porcher-Guinet V, Stagnara C, et al. Altered sleep architecture during the first months of life in infants born to depressed mothers. Sleep Med. 2017;30:195–203.
Halal CS, Bassani DG, Santos IS, Tovo-Rodrigues L, Del-Ponte B, Silveira MF, et al. Maternal perinatal depression and infant sleep problems at 1 year of age: Subjective and actigraphy data from a population-based birth cohort study. J Sleep Res. 2020;:e13047.
Shapiro AF, Jolley SN, Hildebrandt U, Spieker SJ. The effects of early postpartum depression on infant temperament. Early Child Dev Care. 2020;190:1918–30.
Škodová Z, Kelčíková S, Maskálová E, Mazúchová L. Infant sleep and temperament characteristics in association with maternal postpartum depression. Midwifery. 2022;105:103232.
Fernandes DV, Canavarro MC, Moreira H. The mediating role of parenting stress in the relationship between anxious and depressive symptomatology, mothers’ perception of infant temperament, and mindful parenting during the postpartum period. Mindfulness. 2021;12:275–90.
Whiffen VE, Gotlib IH. Infants of postpartum depressed mothers: temperament and cognitive status. J Abnorm Psychol. 1989;98:274.
Scher A, Tirosh E, Lavie P. The relationship between sleep and temperament revisited: evidence for 12-month-olds: a research note. J Child Psychol Psychiatry. 1998;39:785–8.
Nakić Radoš S, Hairston I, Handelzalts JE. The concept analysis of parent-infant bonding during pregnancy and infancy: a systematic review and meta-synthesis. J Reprod Infant Psychol. 2023;:1–24.
Tichelman E, Westerneng M, Witteveen AB, Van Baar AL, Van Der Horst HE, De Jonge A, et al. Correlates of prenatal and postnatal mother-to-infant bonding quality: A systematic review. PloS One. 2019;14:e0222998.
Davies SM, Silverio SA, Christiansen P, Fallon V. Maternal-infant bonding and perceptions of infant temperament: the mediating role of maternal mental health. J Affect Disord. 2021;282:1323–9.
Takács L, Smolík F, Kaźmierczak M, Putnam SP. Early infant temperament shapes the nature of mother-infant bonding in the first postpartum year. Infant Behav Dev. 2020;58:101428.
Dix T. Attributing dispositions to children: An interactional analysis of attribution in socialization. Pers Soc Psychol Bull. 1993;19:633–43.
Murphey DA. Constructing the child: Relations between parents’ beliefs and child outcomes. Dev Rev. 1992;12:199–232.
Parsons JE, Adler TF, Kaczala CM. Socialization of achievement attitudes and beliefs: Parental influences. Child Dev. 1982;:310–21.
Seifer R, Sameroff AJ, Dickstein S, Hayden LC, Schiller M. Family functioning, infant/toddler outcomes, and parental psychopathology. Infant Behav Dev. 1996;19:213.
Putnam SP, Helbig AL, Gartstein MA, Rothbart MK, Leerkes E. Development and assessment of short and very short forms of the Infant Behavior Questionnaire–Revised. J Pers Assess. 2014;96:445–58.
Knafo A. The Longitudinal Israeli Study of Twins (LIST): Children’s social development as influenced by genetics, abilities, and socialization. Twin Res Hum Genet. 2006;9:791–8.
Peterson ER, Mohal J, Waldie KE, Reese E, Atatoa Carr PE, Grant CC, et al. A cross-cultural analysis of the infant behavior questionnaire very short form: an item response theory analysis of infant temperament in New Zealand. J Pers Assess. 2017;99:574–84.
Gartstein MA, Knyazev GG, Slobodskaya HR. Cross-cultural differences in the structure of infant temperament: United States of America (US) and Russia. Infant Behav Dev. 2005;28:54–61.
Buysse DJ, Reynolds CF, Monk TH, Berman SR, Kupfer DJ. The Pittsburgh Sleep Quality Index: a new instrument for psychiatric practice and research. Psychiatry Res. 1989;28:193–213.
Shochat T, Tzischinsky O, Oksenberg A, Peled R. Validation of the Pittsburgh Sleep Quality Index Hebrew translation (PSQI-H) in a sleep clinic sample. Isr Med Assoc J. 2007;9:853.
Brockington IF, Fraser C, Wilson D. The Postpartum Bonding Questionnaire: a validation. Arch Womens Ment Health. 2006;9:233–42.
Brislin RW. Research instruments. Field Methods Cross-Cult Res. 1986;:159–62.
Hairston IS, Solnik-Menilo T, Deviri D, Handelzalts JE. Maternal depressed mood moderates the impact of infant sleep on mother–infant bonding. Arch Womens Ment Health. 2016;:1–11.
Hairston IS, Handelzalts JE, Lehman-Inbar T, Kovo M. Mother-infant bonding is not associated with feeding type: a community study sample. BMC Pregnancy Childbirth. 2019;19:125.
Cox JL, Holden JM, Sagovsky R. Detection of postnatal depression. Development of the 10-item Edinburgh Postnatal Depression Scale. Br J Psychiatry. 1987;150:782–6.
Cox JL, Chapman G, Murray D, Jones P. Validation of the Edinburgh Postnatal Depression Scale (EPDS) in non-postnatal women. J Affect Disord. 1996;39:185–9.
Glasser S, Barell V, Boyko V, Ziv A, Lusky A, Shoham A, et al. Postpartum depression in an Israeli cohort: Demographic, psychosocial and medical risk if actors. J Psychosom Obstet Gynecol. 2000;21:99–108.
Forman DR, O’Hara MW, Larsen K, Coy KC, Gorman LL, Stuart S. Infant emotionality: Observational methods and the validity of maternal reports. Infancy. 2003;4:541–65.
Sadeh A. Iii. Sleep assessment methods. Monogr Soc Res Child Dev. 2015;80:33–48.
Dagla M, Dagla C, Mrvoljak-Theodoropoulou I, Kavakou A-T, Rigoutsou E, Antoniou E. Infant Sleep Difficulties at the 6th Week and the 12th Month Postpartum: What Is their Relationship with Maternal Mental Health and Other Perinatal Factors? Mater Socio-Medica. 2021;33:21.
Dias CC, Figueiredo B. Mother’s prenatal and postpartum depression symptoms and infant’s sleep problems at 6 months. Infant Ment Health J. 2020;41:614–27.
Nolvi S, Karlsson L, Bridgett DJ, Pajulo M, Tolvanen M, Karlsson H. Maternal postnatal psychiatric symptoms and infant temperament affect early mother-infant bonding. Infant Behav Dev. 2016;43:13–23.
Pauli-Pott U, Mertesacker B, Beckmann D. Predicting the development of infant emotionality from maternal characteristics. Dev Psychopathol. 2004;16:19–42.

Table 1: Demographic Characteristics of the Sample
Variable		Statistic
Mothers’ age (M, range)		31.2, 22-46
Education
	High school	8.6%
	Bachelors	45.7%
	MA/PhD/Other	40.6%
	Missing/refused	5.1%
Married		96.2 %
Majority ethnicity		95.2%
Income relative to median
	significantly below	9.6%
	below	16.8%
	around median	30.5%
	above	26.4%
	significantly above	9.1%
	Missing/refused	3.1%
Infants % female		56.5%
Infant age (M, range) [T0]		6.5, 4.4-8.3
Infant age (M, range) [T1]		12.2, 11.6-14
*Note*: N=75

Table 2: Descriptive Statistics

		T0 N=195		T1 N=76	R
mothers’ sleep	PSQI global	6.95 ± 3.65		6.27 ± 3.49	.632**
	above cutoff (>5; %)	56.6%		48.7%
PPD	Sum EPDS scale	5.99 ± 4.5
	above cutoff (>12; %)	7.9
Bonding	PBQ global scale	10.54 ± 9.8		10.63 ± 8.2	.720**
Temperament	Negative affect factor (IBQ-F1)			4.41 ± 0.9
	Positive affect factor (IBQ-F2)			5.63 ± 0.6
	Stranger anxiety factor (IBQ-F3)			4.42 ± 1.2
Infant sleep	TST (minutes)	573.60 ± 99.9		613.03 ± 71.1••	.540**
	WASO (minutes)	63.2 ± 65.5		40.66 ± 43.7•	.185
	# arousals	2.61±2.2		2.17±1.7•	.481**
	Daytime sleep (minutes)		177.60 ± 64.3	146.05 ± 39.6••	.339*
	Nighttime problems factor (ISQ-F1)		0.21 ± 0.2	0.28 ± 0.3•	.424**
	Bedtime resistance factors (ISQ-F2)		0.65 ± 0.4	0.55 ± 0.4•	.628**
Overall sleep difficulty item			3.87 ± 2.4	3.51 ± 2.4	.620**

Note: PSQI = Pittsburgh sleep quality index; EPDS = Edinburgh postpartum depression scale; PBQ = postpartum bonding questionnaire; IBQ = infant behavior questionnaire; ISQ = infant sleep questionnaire; TST=total sleep time, from bedtime to waketime minus time awake at night; WASO=wake after sleep onset. ‘••’=p<.001, ‘•’=p<.050, statistically significant difference between T0 and T1; ‘**’= p<.001, ‘*’= p<.050 significance of Pearson correlation .

Table 3: Hierarchical regression predicting perceived infant sleep problems item at 12 months.

Variable	B	SE	β	CI95%		VIF	R²	ΔR²
				Lower	Upper
Step 1							0.305**
Other kids	-0.77	0.44	-0.179	-1.65	0.11	1.02
Baby TST T1	-0.01	0.004	-0.270**	-0.02	-0.002	1.03
# arousals	0.57	0.15	0.394***	0.27	0.86	1.02

Step 2							0.416***	0.111**
Other kids	-0.70	0.41	-0.163	-1.52	0.11	1.02
Baby TST T1	-0.01	0.004	-0.291**	-0.02	-0.004	1.05
Baby # wake	0.46	0.14	0.318**	0.18	0.74	1.07
PBQ T1	0.06	0.05	0.117	-0.04	0.17	1.08
PSQI T1	0.21	0.07	0.295**	0.07	0.35	1.11

Step 3							0.444***	0.028
Other kids	-0.66	0.42	-0.153	-1.50	0.17	1.08
Baby TST T1	-0.01	0.004	-0.329**	-0.02	-0.005	1.11
Baby # wake	0.46	0.14	0.322**	0.19	0.74	1.08
PBQ T1	0.02	0.06	0.037	-0.09	0.13	1.31
PSQI T1	0.20	0.08	0.275*	0.02	0.37	1.83
EPDS T0	1.20	0.69	0.222	-0.17	2.58	1.85
PSQI T0	-0.04	0.09	-0.063	-0.23	0.15	2.28

Note: N=72; CI=confidence interval; VIF=variance inflation factor; TST=total sleep time; # arousals = number of nighttime awakenings; PSQI = Pittsburgh sleep quality index; EPDS = Edinburgh postpartum depression scale; PBQ = postpartum bonding questionnaire; ***=p<.001; **=p<.01, *=p<.05.

Table 4: Hierarchical regression predicting the negative affect factor of IBQ at 12 months.

Variable	B	SE	β	CI95%		VIF	R²	ΔR²
				Lower	Upper
Step 1							0.083
Other kids	0.39	0.19	0.246*	0.02	0.77	1.02
Baby TST T1	0.00	0.002	0.062	-0.002	0.004	1.02
# arousals	0.08	0.06	0.149	-0.05	0.20	1.02

Step 2							0.162*	0.084*
Other kids	0.41	0.18	0.255*	0.04	0.77	1.02
Baby TST T1	0.00	0.002	0.023	-0.003	0.003	1.05
# arousals	0.06	0.06	0.106	-0.07	0.18	1.03
PBQ T1	0.05	0.23	0.266*	-0.006	0.10	1.08
PSQI T1	0.02	0.03	0.067	-0.05	0.08	1.11

Step 3							0.224*	0.061
Other kids	0.48	0.18	0.301*	0.11	0.85
Baby TST T1	0.00	0.002	-0.036	-0.003	0.003	1.07
# arousals	0.05	0.06	0.100	-0.07	0.18	1.11
PBQ T1	0.03	0.03	0.167	-0.02	0.08	1.07
PSQI T1	0.04	0.04	0.170	-0.03	0.12	1.30
EPDS T0	0.63	0.30	0.311*	0.02	1.23	1.82
PSQI T0	-0.07	0.04	-0.291	-0.16	0.01	1.83

Note: IBQ=Infant Behavior Questionnaire; CI=confidence interval; VIF=variance inflation factor; # arousals = number of nighttime awakenings; PSQI = Pittsburgh sleep quality index; EPDS = Edinburgh postpartum depression scale; PBQ = postpartum bonding questionnaire; ***=p<.001; **=p<.01, *=p<.05.

No competing interests reported.

supplement.docx

Associations between maternal characteristics, infant sleep, and perceived infant temperament in the first year of life: A longitudinal study

Status:

Version 1

Abstract

Figures

Introduction

Methods

Participants

Analytical methods

Results

Discussion

Conclusions

Declarations

References

Tables

Additional Declarations

Supplementary Files

Status:

Version 1